Pre-mRNA splicing must occur with high fidelity to prevent catastrophic errors. Yet, the molecular mechanisms of fidelity in splicing are understood poorly. Splicing is catalyzed by the spliceosome, a dynamic ribonucleoprotein machine in which small nuclear RNA (snRNA) components play key roles in substrate recognition and catalysis. The long-term goals of this proposal are to understand how fidelity is established in splicing and how spliceosome dynamics promote fidelity. We aim to understand how the spliceosome promotes splicing of a genuine substrate and antagonizes splicing of an aberrant substrate. Further, we aim to understand how splicing of a genuine substrate and stalling of an aberrant substrate each lead to spliceosome disassembly. Recently, we have made significant breakthroughs in understanding how the U2 and U6 snRNAs promote splicing of a genuine substrate, how the DExD/H box ATPase Prp22 antagonizes splicing of an aberrant substrate and how spliceosome disassembly is regulated. Specifically, we aim (i) to determine the role of Prp22 in antagonizing aberrant intermediates, (ii) to determine the roles of the DExD/H box ATPases Prp43 and Brr2 and the EF-2-like GTPase Snu114 in spliceosome disassembly, (iii) to investigate the role of U2/U6 helix I sequences in promoting splicing and (iv) to determine the role of the U2 loop Ha in remodeling the spliceosome for exon ligation. We will pursue these aims using a combined approach of molecular genetics and biochemistry. These studies will likely have broad implications for understanding (i) the function and regulation of DExD/H box ATPases, (ii) the mechanisms for establishing fidelity in splicing and (iii) the mechanisms for regulating the activity of the spliceosome. Further, as fidelity is intimately linked to splice site choice, these studies promise to impact our understanding of the mechanisms for regulating splicing. To investigate splicing, we will utilize the model organism baker's yeast, which resembles a simplified human cell and which has served as a long-standing model for human disease. Significantly, as life requires the faithful expression of genes, splicing errors commonly lead to human disease. Thus, these studies will lead to an understanding of how fidelity is established in human cells, how compromised fidelity may contribute to disease and how fidelity might be engineered to treat disease.